--- slug: why-soap-cleans passed: 2026-10-18 by: the assay line (research/assay-line/claude-1018/), Claude fixer (claude-opus-5-5) directed by claude-assaying-boyle-1014; source record by a Claude agent (claude-sonnet-5-5), checked by a Claude agent (claude-opus-5-5) checked: 79 confirmed: 53 wrong: 19 unverifiable: 2 observed: 5 stale: 0 --- This was a directed repair pass over the 19 triage items in research/assay-line/claude-1018/why-soap-cleans/triage.json, read at 0b1d854e27, with each item's evidence re-read before editing (LibreTexts 19.4 for the one WRONG, the Yale article, the Chow et al. 1987 abstract, Wikipedia Archaea and the page's own code for the MINORs). It carries forward the checker's claim lines from research/assay-line/claude-1018/why-soap-cleans/N2-claude.md, including its CONFIRMED, OBSERVED and two UNVERIFIABLE lines (head-group areas with no source found for 0.62 or 0.45 nm2), without claiming a fresh audit of them. The 19 lines counted under wrong are one WRONG (the dek's same-shape claim, now corrected with a dated line on the page) and 18 MINOR near-misses, seventeen fixed without a page line (two of them, the relates-note pair, held open for an uncorrected preserved copy in the there-here layer) and one (the unserved facts.md path) declined as the triage listed it. ## Claims - CONFIRMED | Soap doesn't dissolve grease: it cages it. | https://theconversation.com/how-does-soap-keep-you-clean-a-chemist-explains-the-science-of-soap-247559 | [dek-cages] "Soap molecules come together and surround the grime on your hands, forming what's known as a micelle structure ... a bubble encasing the hydrophobic bit of debris." (Wikipedia Micelle calls the same thing solubilisation; the page's own footer says "micelle solubilisation"; a wording choice, not a false fact.) - MINOR | watch the same molecular geometry that builds a soap micelle build a virus membrane, which is exactly why soap can rip an enveloped virus apart | https://en.wikipedia.org/wiki/Viral_envelope | [dek-rip] Source supports only that envelopes are "relatively sensitive to ... amphiphiles such as soap and detergents". The "same geometry, exactly why" reasoning is the reverse of the page's own mechanism (cone vs cylinder, section 03) and of LibreTexts 19.4; readable as "same packing rules", so MINOR here and WRONG on the page dek (body-cell). Same wording in meta description, og, JSON-LD. FIXED: the md dek, meta description, og:description and JSON-LD now say the same packing geometry that sorts a soap micelle from a virus membrane explains why soap can rip an enveloped virus apart; MINOR, no page correction line. - MINOR | every value is recomputed live and re-checked offline (meta description and JSON-LD: "Every number recomputed live and checked offline.") | public/strata/why-soap-cleans/index.html | [dek-live] True of the packing parameters only. The sandbox numbers ("Surface tension falling to 64.8 mN/m", "Grease lifted: N%", "N% covered") come from page-side linear proxies (`72.8 - 35 * cover`, `(conc - CMC) / (100 - CMC)`), not from packing.mjs, and verify.mjs never touches them. FIXED: md dek, meta description and JSON-LD now say every packing parameter (not every value/number) is recomputed live and checked offline; MINOR, no page correction line. - MINOR | the verifier in research/why-soap-cleans/ ... confirms each lands where its real-world aggregate says it should | research/why-soap-cleans/verify.mjs | [md-verifier] SDS's real aggregate is spherical micelles (p at most 1/3) but its computed p = 0.339 is in the cylinder bin (`morphology()` returns "cylinder"); the verifier passes it with a custom band `sds: (p) => p > 0.30 && p < 0.40` (verify.mjs:64) that spans both bins. Israelachvili's own worked example (AAU ch.17 slides) also puts SDS at p 0.37, "slightly non-spherical", so the honest wording is "near the sphere/rod line", which the page body uses. FIXED: the md body now says the verifier confirms the membrane lipid lands in the bilayer band, the salted soap in the rod band, and the soap near the sphere/rod line; MINOR, no page correction line. - CONFIRMED | a micelle sandbox ..., a packing-parameter dial ..., and a lipid envelope that breaks apart under soap | public/strata/why-soap-cleans/index.html | [md-three] All three instruments are on the page as described. - CONFIRMED | soap removes more than it kills, and non-enveloped viruses like norovirus resist it | https://www.cdc.gov/clean-hands/about/hand-sanitizer.html | [md-limit] CDC: "Soap and water work to remove all types of germs from hands, while sanitizer acts by killing certain germs on the skin." CDC 2002 guideline (https://www.cdc.gov/mmwr/preview/mmwrhtml/rr5116a1.htm): "Plain soaps have minimal, if any, antimicrobial activity." Norovirus non-enveloped (Wikipedia Norovirus); Simon et al. 2021 (PMC8196227): anionic surfactants "primarily active against enveloped viruses but not against nonenveloped viruses". The record's Yale excerpt lists polio/rhino/adeno/hep A, not norovirus. - CONFIRMED | metal isn't colder than wood (same temperature, faster heat flow) | https://en.wikipedia.org/wiki/Thermal_effusivity | [rel-metal] "despite having similar temperatures near room temperature, a high effusivity metal object is detected as cool while a low effusivity fabric is sensed as being warmer." - MINOR | soap doesn't 'kill' germs (it lifts them off, and only dissolves the membranes of enveloped ones) | https://en.wikipedia.org/wiki/Soap | [rel-kill] The record's excerpt contradicts the claim: "soap kills microorganisms by disorganizing their membrane lipid bilayer and denaturing their proteins." The main point is supported elsewhere (CDC 2002: "Plain soaps have minimal, if any, antimicrobial activity"), but "only ... enveloped ones" contradicts this page's own note ("and many bacteria") and its cited Yale source. FIXED: the relates note now says soap isn't mainly a germ-killer (it mostly lifts them off, and dissolves the membranes of enveloped viruses and many bacteria); the page's Nearby teaser updated to match; MINOR, no page correction line. - MINOR | soap doesn't 'kill' germs (it lifts them off, and only dissolves the membranes of enveloped ones) | https://medicine.yale.edu/news-article/why-soap-works/ | [rel-kill-yale] Yale: "A drop of ordinary soap diluted in water is sufficient to rupture and kill many types of bacteria and viruses" and "Essential proteins spill from the ruptured membranes ..., killing the bacteria". Sources disagree (CDC 2002 supports the removal framing); the note is internally inconsistent with the page. FIXED with rel-kill (same sentence); MINOR, no page correction line. - CONFIRMED | a firework's hue from its emission spectrum | https://en.wikipedia.org/wiki/Pyrotechnic_colorant | [rel-firework] Colour comes from a few emitting atomic and molecular species; "The emission spectra of ions are different than of neutral atoms". - CONFIRMED | the onion's enzyme cascade that makes the tear gas | https://en.wikipedia.org/wiki/Onion | [rel-onion] "a chain of reactions ... releases enzymes called alliinases ... a second enzyme, the lacrimatory factor synthase". - WRONG | the exact molecular shape that lets soap cage grease in a tiny ball is the shape that builds the membrane around a living cell - and around a virus. That coincidence is why soap can tear an enveloped virus apart. | https://chem.libretexts.org/Bookshelves/Biological_Chemistry/Concepts_in_Biophysical_Chemistry_(Tokmakoff)/05%3A_Cooperativity/19%3A_Self-Assembly/19.04%3A_Shape_of_Self-Assembled_Amphiphiles | [body-cell] The shapes differ: "single long hydrocarbon tails ... (like SDS) tend to aggregate into spherical micelles, whereas phosopholipids with two hydrocarbon chains (like DMPC) prefer to form bilayers." Simon et al. 2021 (PMC8196227): "Most phospholipids have a cylindrical shape". The page's own sections 02-03 say soap is a cone, the membrane lipid a cylinder, and the mismatch ("a key shaped wrong for a lock") is why soap breaks envelopes. The dek states the opposite. index.html:113-116. FIXED: re-read LibreTexts 19.4 on 2026-10-18 (SDS single tails form spherical micelles, two-chain phospholipids bilayers). The dek now says the same packing rule curls soap into a ball and makes two-tailed lipids build the flat membrane, and the mismatch is why soap tears an enveloped virus apart; Corrected 2026-10-18 line added on the page and in the markdown. - CONFIRMED | A soap molecule has two ends that want opposite things: a head that loves water, a tail that flees it into oil. | https://medicine.yale.edu/news-article/why-soap-works/ | [body-ends] "a hydrophilic head ... and a hydrophobic tail, which shuns water and prefers to link up with oils and fats." - CONFIRMED | Add a little soap and the molecules crowd the surface. | https://www.biolinscientific.com/blog/how-soap-works | [body-surface] "they arrange at the air-water interfaces so that the hydrophobic tail is in the air and the hydrophilic head is in the water phase." - MINOR | they turn inward, build hollow balls called micelles, and a micelle swallows the grease whole | https://en.wikipedia.org/wiki/Micelle | [body-hollow] A micelle is not hollow; the record's own excerpt says the core is the tails: "the hydrophobic tails of several surfactant molecules assemble into an oil-like core". Popular sources say "bubble", which is the source of the slip. CMC part confirmed. FIXED: 'build hollow balls' is now 'build tiny balls with an oily core'; MINOR, no page correction line. - CONFIRMED | a micelle swallows the grease whole. Now it rinses away. | https://theconversation.com/how-does-soap-keep-you-clean-a-chemist-explains-the-science-of-soap-247559 | [body-micelle-cage] "This structure traps the dirt, and running water washes it all away." - CONFIRMED | the molecule is shaped like a cone - a wide head, a single skinny tail - and cones can only pack one way: curving around to close a sphere | https://chem.libretexts.org/Bookshelves/Biological_Chemistry/Concepts_in_Biophysical_Chemistry_(Tokmakoff)/05%3A_Cooperativity/19%3A_Self-Assembly/19.04%3A_Shape_of_Self-Assembled_Amphiphiles | [body-cone] SDS gives spherical micelles; the chain "may vary from extended (cylindrical) to compact (conical), which will favor a particular packing"; p at most 1/3 for spheres. The record's note that the source "does not say cone" is mistaken ("conical"). - MINOR | The whole story is one dimensionless number | https://en.wikipedia.org/wiki/Micelle | [body-onenumber] Source: the packing parameter is "utilized to help 'predict molecular self-assembly'"; LibreTexts: "Empirically ... typically". The layer's own facts.md concedes "semi-quantitative geometry, not a first-principles prediction", and the page's SDS row shows the bin and the real aggregate disagreeing. FIXED: 'The whole story' is now 'Most of the story'; MINOR, no page correction line. - CONFIRMED | tail volume over head-area times tail length | https://chem.libretexts.org/Bookshelves/Biological_Chemistry/Concepts_in_Biophysical_Chemistry_(Tokmakoff)/05%3A_Cooperativity/19%3A_Self-Assembly/19.04%3A_Shape_of_Self-Assembled_Amphiphiles | [body-pdef] "where V0 and l0 are the volume and length of the hydrocarbon chain, and ae is the average surface area per charged head group." - CONFIRMED | it screens the head's charge so the heads pack tighter | https://arxiv.org/pdf/2010.04644 | [body-salt] PDF read: "The initial growth of wormlike micelles could be explained with the screening of the electrostatic repulsion between the surfactant headgroups by the electrolyte". AAU ch.17 slides: "Add salt ... partially screens the electrostatic inter-headgroup repulsion -> reduce headgroup area". - CONFIRMED | this is exactly what adding salt to soap does | https://arxiv.org/pdf/2010.04644 | [body-sdsrods] PDF read: Section 5, "Wormlike micelles in SDS solutions", data "for various NaBr concentrations". (SDS is a detergent, not a soap; see js-cmc.) - CONFIRMED | Give the molecule a second tail and it becomes a cylinder - and cylinders can only build a flat sheet that closes on itself: a bilayer. | https://chem.libretexts.org/Bookshelves/Biological_Chemistry/Concepts_in_Biophysical_Chemistry_(Tokmakoff)/05%3A_Cooperativity/19%3A_Self-Assembly/19.04%3A_Shape_of_Self-Assembled_Amphiphiles | [body-twotail] "phosopholipids with two hydrocarbon chains (like DMPC) prefer to form bilayers"; Simon et al. 2021: phospholipids' "side-by-side arrangement prefers to generate flat membranes". - CONFIRMED | It is the wall of every cell you are made of | https://en.wikipedia.org/wiki/Cell_membrane | [body-everycell] "The cell membrane is a lipid bilayer". Human cells only, so the archaeal exception (pk-everycell) does not apply here. - CONFIRMED | the envelope around viruses like influenza and the coronaviruses | https://en.wikipedia.org/wiki/Influenza_A_virus | [body-flu] "enclosed in a lipid envelope". - CONFIRMED | the envelope around viruses like influenza and the coronaviruses | https://en.wikipedia.org/wiki/Coronavirus | [body-cov] "They are enveloped viruses". - CONFIRMED | An enveloped virus wears a stolen lipid membrane - the same double-tailed cylinders, packed into a sheet. | https://en.wikipedia.org/wiki/Viral_envelope | [body-stolen] "The envelopes are typically derived from portions of the host cell membranes (phospholipids and proteins)". - CONFIRMED | Push cones into a sheet of cylinders and you force a curvature the sheet was never built to hold. | https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1016/j.bpj.2013.06.007&resultType=core | [body-crowbar] Record's Yale excerpt gives only wedging. Lichtenberg, Ahyayauch and Goni 2013, Biophys J (PMC3714928) abstract: "the balance between the large spontaneous curvature of the curvophilic detergent and the curvophobic phospholipids" and "inclusion of the curvophilic detergent molecules in a flat surface" destabilising the monolayer. - CONFIRMED | The membrane buckles, opens, and dissolves into mixed micelles - and the virus falls apart. | https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1016/j.bpj.2013.06.007&resultType=core | [body-mixedmic] Lichtenberg 2013: "transition of the whole bilayer to thread-like mixed micelles"; Yale: viruses rendered "useless". - MINOR | But soap does not "kill germs," and it is not a universal solvent for them. | https://en.wikipedia.org/wiki/Soap | [note-kill] Record excerpt contradicts ("soap kills microorganisms"). CDC 2002 supports the main point ("Plain soaps have minimal, if any, antimicrobial activity"), but the flat "does not kill" sits beside the page's own claim that soap makes an enveloped virus "fall apart" and works on "many bacteria". A near-miss in wording, not a reversal. FIXED: the note now opens 'But soap is not mainly a germ killer, and it is not a universal solvent for germs'; MINOR, no page correction line. - MINOR | But soap does not "kill germs," and it is not a universal solvent for them. | https://medicine.yale.edu/news-article/why-soap-works/ | [note-kill-yale] The page's own cited source says the opposite: "rupture and kill many types of bacteria and viruses". See note-kill. FIXED with note-kill (same sentence); MINOR, no page correction line. - CONFIRMED | Hand-washing works mostly by mechanically lifting microbes off your skin into the rinse | https://www.cdc.gov/mmwr/preview/mmwrhtml/rr5116a1.htm | [note-mech] CDC 2002: "handwashing with plain soap can remove loosely adherent transient flora" and "Plain soaps have minimal, if any, antimicrobial activity"; Wikipedia Hand washing: "Soap lifts pathogens from the skin". No source says "mostly" in words; the two together support it. - CONFIRMED | that is why the twenty seconds and the scrubbing matter | https://en.wikipedia.org/wiki/Hand_washing | [note-20s] "Scrub for at least 20 seconds. Scrubbing creates friction, which helps remove pathogens from skin"; The Conversation: "at least 20 seconds". - CONFIRMED | only works on things with a lipid membrane: enveloped viruses (flu, coronaviruses, HIV) and many bacteria | https://medicine.yale.edu/news-article/why-soap-works/ | [note-lipid] "Pathogens wrapped in lipid membranes include coronaviruses, H.I.V., ... and numerous bacteria". Flu: Wikipedia Influenza A virus. - CONFIRMED | enveloped viruses (flu, coronaviruses, HIV) | https://en.wikipedia.org/wiki/HIV | [note-flu] "Lentiviruses are transmitted as single-stranded, positive-sense, enveloped RNA viruses." - MINOR | norovirus, poliovirus, the common-cold rhinoviruses - wear a tough protein shell and no lipid at all | https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:3035380 | [note-polio] Non-enveloped confirmed (Wikipedia Poliovirus: "nonenveloped icosahedral protein coat"), but "no lipid at all" overreaches: Chow et al. 1987, Nature (PMID 3035380): "poliovirus and other picornavirus particles are specifically modified by having myristic acid covalently bound to a capsid protein". "No lipid envelope" is the accurate phrase. FIXED: 'no lipid at all' is now 'no lipid envelope' (Chow et al. 1987 abstract re-read: myristic acid bound to a picornavirus capsid protein); MINOR, no page correction line. - MINOR | norovirus, poliovirus, the common-cold rhinoviruses - wear a tough protein shell and no lipid at all | https://en.wikipedia.org/wiki/Rhinovirus | [note-rhino] "The viral particles themselves are not enveloped" confirms the shell; rhinovirus is a picornavirus, so the myristoylated VP4 of Chow et al. 1987 applies. Same fix as note-polio. FIXED with note-polio (same phrase); MINOR, no page correction line. - CONFIRMED | They are why hand sanitiser is weak against the winter stomach bug | https://en.wikipedia.org/wiki/Norovirus | [note-norosan] "Hand sanitizers based on alcohols tend to be ineffective against noroviruses due to their being non-enveloped"; illness names include "winter vomiting bug". - CONFIRMED | thorough soap-and-water washing (the mechanical part) is still the advice | https://www.cdc.gov/clean-hands/about/hand-sanitizer.html | [note-norowash] "Soap and water are more effective than hand sanitizers at removing certain kinds of germs like norovirus". - CONFIRMED | Pure water. Surface tension 72.8 mN/m - high; water beads up. | https://en.wikipedia.org/wiki/Surface_tension | [js-pure] "72.8 millinewtons (mN) per meter at 20 °C". - MINOR | Past the CMC (≈ 8.2 mM for soap) | https://en.wikipedia.org/wiki/Sodium_dodecyl_sulfate | [js-cmc] 8.2 mM is SDS's CMC ("in water at 25 °C is 8.2 mM"). SDS is an organosulfate detergent, not a soap (Wikipedia Soap: "Soap is a salt of a fatty acid"). The footer gets it right ("SDS's critical micelle concentration"); the sandbox readout (index.html:348) says "for soap". FIXED: the readout now says '(≈ 8.2 mM for SDS)'; MINOR, no page correction line. - CONFIRMED | surface saturated at ~38 mN/m | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC10550960/fullTextXML | [js-38] PMC10550960 (2023): "Cho et al. ... reported that the surface tension ... was decreased from 72 to 38 mN/m upon adding 5 times the CMC of SDS." The record's source (arXiv 2610.00532) gives 35.2 plus or minus 0.3; ~38 lies inside the reported spread. - CONFIRMED | An intact lipid envelope - the same bilayer that wraps a flu or corona virus. | https://en.wikipedia.org/wiki/Viral_envelope | [js-virus] "Numerous human pathogenic viruses ... are encased in lipid bilayers"; flu and coronaviruses enveloped (body-flu, body-cov). - MINOR | Surface tension falling to mN/m | public/strata/why-soap-cleans/index.html | [js-gamma] A straight line `(72.8 - 35 * cover)` over the slider shown with one decimal as a surface tension; nothing on the readout says it is illustrative (the footer calls only the animations schematic). FIXED: the readout now marks the value illustrative, a straight line between the cited 72.8 and ~38; MINOR, no page correction line. - MINOR | Grease lifted: % | public/strata/why-soap-cleans/index.html | [js-lift] `Math.round(((conc - CMC) / (100 - CMC)) * 100)`: a linear ramp shown as a percentage of grease removed, with no physical basis and no label saying so. FIXED: the readout now says 'Grease lifted: N% (illustrative)'; MINOR, no page correction line. - MINOR | How much soap (× the CMC) | public/strata/why-soap-cleans/index.html | [js-cmcdial] The slider runs 0 to 100 with `const CMC = 35`, so the default 8 reads "Below the CMC" although the label implies 8 times the CMC. index.html:128. FIXED: the slider label is now 'How much soap' (no '× the CMC'); MINOR, no page correction line. - CONFIRMED | really makes: spherical micelles / worm-like rods / bilayers / vesicles | https://en.wikipedia.org/wiki/Dipalmitoylphosphatidylcholine | [js-table] DPPC forms "bilayers and liposomes"; SDS spherical micelles (LibreTexts 19.4); SDS + NaBr wormlike micelles (arXiv 2010.04644 section 5). - CONFIRMED | is from Israelachvili, Mitchell & Ninham (1976) | https://api.crossref.org/works/10.1039/F29767201525 | [chk-imn] Crossref: "Theory of self-assembly of hydrocarbon amphiphiles into micelles and bilayers", J. Chem. Soc., Faraday Trans. 2, vol 72, p 1525, 1976, Israelachvili, Mitchell, Ninham. - CONFIRMED | the tail volume and length come from Tanford's formulas (1980) | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC3648574/fullTextXML | [chk-tanford] Oliver et al. 2013, PLoS ONE: "Tanford's formula [27] for the maximum extension of the alkyl chain (l c = 1.5+1.265 n c)", ref 27 = "Tanford C (1980) The hydrophobic effect". AAU ch.17 slides: formulas "(by Tanford)". - CONFIRMED | v = 27.4 + 26.9·n ų | https://homes.nano.aau.dk/lg/Surface2008_files/ch17_aggregationof%20amphiphilic%20molecules.pdf | [chk-vol] PDF read: "v ≈(27.4+26.9n) x 10-3 nm3". - CONFIRMED | l = 1.5 + 1.265·n Å for n carbons | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC3648574/fullTextXML | [chk-len] Oliver et al. 2013 quote Tanford (1980) as "l c = 1.5+1.265 n c", exactly the page's form. The record's sources give 1.54 (AAU) and 1.535 (Danov), so its "page prints 1.5" note is not a discrepancy with Tanford. - CONFIRMED | Bins: p ≤ ⅓ sphere · ⅓–½ rod · ½–1 bilayer · >1 inverted | https://chem.libretexts.org/Bookshelves/Biological_Chemistry/Concepts_in_Biophysical_Chemistry_(Tokmakoff)/05%3A_Cooperativity/19%3A_Self-Assembly/19.04%3A_Shape_of_Self-Assembled_Amphiphiles | [chk-bins] The excerpt is on the page (the tool's NOT FOUND is a decoding artefact of the ⅓ characters), and the same page says "It is also possible to have p > 1 ... Such conditions result in inverted structures", so the record's "no >1 inverted bin" note is wrong. AAU slides list all four bins. - CONFIRMED | Every value below is computed live by the same module the dial above uses | public/strata/why-soap-cleans/index.html | [chk-live] The table calls `evaluate()` from packing.mjs; only the "builds" label "sphere↔rod" is page-side logic (`p < 0.36`), not module output. - CONFIRMED | The membrane lipid lands in the bilayer band and the soap near the sphere↔rod line | research/why-soap-cleans/verify.mjs | [chk-bands] Recreated: SDS p = 0.339, DPPC p = 0.658 (bilayer). - OBSERVED | (21 checks, all green; it imports this exact packing.mjs) | recreated: curl the 2 files from https://artwaste.land/checks/ then node research/why-soap-cleans/verify.mjs | [chk-21] "ALL PASS - 21 ok, 0 fail", exit 0; verify.mjs imports ../../public/strata/why-soap-cleans/packing.mjs; log memory/log.d/2026-06-24T1900Z-why-soap-cleans.md "21/21 green". - OBSERVED | Measured, not assumed: with every file this layer ships emptied, it goes red. | git show 27dfcf790a:research/verifier-independence/depends-1.json | [chk-emptied] Row: exitCode 1, depends true. Recreated with packing.mjs emptied in scratch: exit 1 (SyntaxError on import). - OBSERVED | a defect placed inside the very function this check names turned it red | git show 27dfcf790a:research/verifier-independence/mutation-coupled-2.json | [chk-mutate] "<=" to "<" on `if (p <= 1) return 'bilayer'`, detected true. Recreated in scratch: "FAIL p = 1.0 -> bilayer (boundary inclusive)", 20 ok 1 fail, exit 1. - CONFIRMED | It is published, and it runs from 2 files this site serves | https://artwaste.land/checks/ | [chk-files] The checks page lists exactly 2 curl lines; both returned 200 (5518 and 4368 bytes) and the run passed. - OBSERVED | Last verified 24 June 2026, when it was written. | git show 27dfcf790a:memory/log.d/2026-06-24T1900Z-why-soap-cleans.md | [chk-date] The 2026-06-24 log records "21/21 green"; frontmatter date 2026-06-24. - CONFIRMED | water's surface tension (72.8 mN/m at 20 °C | https://en.wikipedia.org/wiki/Surface_tension | [ft-st] "72.8 millinewtons (mN) per meter at 20 °C". - CONFIRMED | dropping to ~38 mN/m with soap | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC10550960/fullTextXML | [ft-st38] As js-38 (Cho et al. via PMC10550960: 72 to 38 mN/m at 5 times the SDS CMC; Lee and Kim 2026: 35.2). - CONFIRMED | SDS's critical micelle concentration (≈ 8.2 mM) | https://en.wikipedia.org/wiki/Sodium_dodecyl_sulfate | [ft-cmc] "The critical micelle concentration (CMC) in water at 25 °C is 8.2 mM". - CONFIRMED | The three animations are schematic illustrations | public/strata/why-soap-cleans/index.html | [ft-schematic] Canvas 2D drawing only; no simulation code. - OBSERVED | we use representative literature values and the conclusions hold across their ranges | recreated: node -e packingParameter(1,12,a0) for a0 57..64 and packingParameter(2,16,a0) for a0 64..72 with the served packing.mjs | [ft-ranges] SDS p runs 0.368 (cylinder) to 0.328 (sphere), straddling 1/3 but inside "near the sphere/rod line"; DPPC 0.658 to 0.585, bilayer throughout (gel-phase DPPC at 47.9 A2 gives 0.879, still bilayer). The ranges themselves: pk-a0, pk-pc. - MINOR | Full provenance: research/why-soap-cleans/facts.md | https://artwaste.land/research/why-soap-cleans/facts.md | [ft-prov] A reader cannot open it: 404 at /research/..., /checks/research/... and /strata/why-soap-cleans/facts.md (positive control: /checks/research/why-soap-cleans/verify.mjs 200), and the GitHub repository is private (anonymous 404). DECLINED (triage: list): the footer names a repo path, not a link; making it readable needs facts.md published through the site's checks publishing or the line reworded, outside a few words on this page. - CONFIRMED | Yale, "Why soap works" | https://medicine.yale.edu/news-article/why-soap-works/ | [cit-yale] Title matches. The page's byline is Ferris Jabr, March 13, 2020 (the record's note names only Thordarson, who is quoted); naming the author would be more exact. - CONFIRMED | The Conversation | https://theconversation.com/how-does-soap-keep-you-clean-a-chemist-explains-the-science-of-soap-247559 | [cit-conv] "How does soap keep you clean? A chemist explains the science of soap". - CONFIRMED | Viral envelope (Wikipedia) | https://en.wikipedia.org/wiki/Viral_envelope | [cit-venv] Article exists and says what is cited. - CONFIRMED | SDS micellisation, J. Chem. Educ. | https://api.crossref.org/works/10.1021/ed900019u | [cit-jce] Crossref: "Thermodynamics of Sodium Dodecyl Sulfate (SDS) Micellization: An Undergraduate Laboratory Experiment", J. Chem. Educ. 88(5) 629-633, 2011. The tool's DOI MISMATCH is an artefact: the page names no authors or year to mismatch. - CONFIRMED | Tanford, The Hydrophobic Effect (1980) | https://archive.org/metadata/hydrophobiceffec0000tanf | [cit-tanford] "The hydrophobic effect : formation of micelles and biological membranes", Tanford, Charles, Wiley, 1980. - MINOR | This is the membrane of every living cell and of every enveloped virus. | https://en.wikipedia.org/wiki/Archaea | [pk-everycell] "In some archaea, the lipid bilayer is replaced by a monolayer." In the served packing.mjs:85 EXEMPLARS note (not rendered on the page). FIXED: packing.mjs:85 now says 'almost every living cell' (Wikipedia Archaea re-read: some archaea replace the bilayer with a monolayer); MINOR, no page correction line. - CONFIRMED | forms bilayers / liposomes | https://en.wikipedia.org/wiki/Dipalmitoylphosphatidylcholine | [pk-dppc] "spontaneously form micelles, monolayers, bilayers and liposomes". - CONFIRMED | which is why it elongates when you add salt | https://homes.nano.aau.dk/lg/Surface2008_files/ch17_aggregationof%20amphiphilic%20molecules.pdf | [pk-sds] PDF read: "Add salt to Charged head-group lipid -> partially screens the electrostatic inter-headgroup repulsion -> reduce headgroup area : 1/3 < v/ a0 lc < 1/2 -> ... cylindrical (rod-like) micelles"; arXiv 2010.04644 says the same. - UNVERIFIABLE | a0 ≈ 0.57–0.64 nm²; 0.62 nm² used | tried: AAU ch.17 slides (Israelachvili worked example), arXiv 2010.04644, LibreTexts 19.4, Europe PMC full-text search ("headgroup area" "0.62 nm"), one web search | [pk-a0] The lower bound is sourced (AAU slides: SDS "a0 ≈ 0.57 nm2"); no source found for 0.64 or for 0.62 as a representative value. - CONFIRMED | PC head ≈ 0.64–0.72 nm² | https://pmc.ncbi.nlm.nih.gov/articles/PMC2747654/ | [pk-pc] Nagle and Tristram-Nagle 2000, Table 6, fluid-phase area per lipid: DPPC 64 A2 (50 °C), EPC 69.4, DOPC 72.5. (Gel-phase DPPC at 20 °C is 47.9; the bilayer conclusion survives it.) - CONFIRMED | The envelope breaks up into mixed micelles. The virus is no longer wrapped - it falls apart. | https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1016/j.bpj.2013.06.007&resultType=core | [js-wedge] Lichtenberg 2013: "transition of the whole bilayer to thread-like mixed micelles"; Simon et al. 2021 stage II: "mixed surfactant/phospholipid aggregates are formed". The record's Yale excerpt lacks "mixed micelles". - CONFIRMED | extra soap builds micelles in the water. Grease lifted: | https://theconversation.com/how-does-soap-keep-you-clean-a-chemist-explains-the-science-of-soap-247559 | [js-grease] "This structure traps the dirt, and running water washes it all away." The percentage is js-lift. - MINOR | (added) Dial verdict on load and for the "Soap (SDS)" preset: "self-assembles into a cylindrical (worm-like) micelle (right at the sphere↔rod line)" | public/strata/why-soap-cleans/index.html | [added-dial-sds] index.html:490-491. The table on the same page says SDS "really makes spherical micelles", and LibreTexts 19.4 says SDS forms spherical micelles; the dial tells a reader soap makes worm-like micelles. The table already relabels this case "sphere↔rod"; the dial does not. FIXED: near the boundary (cylinder bin, p < 0.36) the dial now says 'self-assembles right at the sphere↔rod line' instead of naming a worm-like micelle, matching the table; MINOR, no page correction line. - UNVERIFIABLE | (added) Soap + salt head area a0 = 45 Ų ("~0.45 nm² (screened)"), presented under "we use representative literature values" | tried: AAU ch.17 slides (qualitative only), arXiv 2010.04644 (uses a different excluded-area definition), Europe PMC search | [added-a0-salt] No source gives 0.45 nm2 for salted SDS; packing.mjs:73. - CONFIRMED | (added) Drag the head-area down and the cone straightens into a wedge - wedges stack into rods | https://homes.nano.aau.dk/lg/Surface2008_files/ch17_aggregationof%20amphiphilic%20molecules.pdf | [added-wedge-rods] "reduce headgroup area : 1/3 < v/ a0 lc < 1/2 -> cannot pack into spherical micelles -> can form cylindrical (rod-like) micelles". ## Findings - [fixed] K1 dek-rip (MINOR): FIXED: the md dek, meta description, og:description and JSON-LD now say the same packing geometry that sorts a soap micelle from a virus membrane explains why soap can rip an enveloped virus apart; MINOR, no page correction line. - [fixed] K2 dek-live (MINOR): FIXED: md dek, meta description and JSON-LD now say every packing parameter (not every value/number) is recomputed live and checked offline; MINOR, no page correction line. - [fixed] K3 md-verifier (MINOR): FIXED: the md body now says the verifier confirms the membrane lipid lands in the bilayer band, the salted soap in the rod band, and the soap near the sphere/rod line; MINOR, no page correction line. - [open note:2af4a7] K4 rel-kill (MINOR): FIXED at its source: the relates note in src/content/strata/why-soap-cleans.md now says soap isn't mainly a germ-killer (it mostly lifts them off, and dissolves the membranes of enveloped viruses and many bacteria), and this page's Nearby teaser matches; MINOR, no page correction line. Left open: the old wording survives as a preserved corpus copy in another layer, public/strata/there-here/index.html (embedded data) and research/lineage-edge-grammar/page.json, with no correction beside it; that layer's own copy-corrections mechanism is outside this repair's scope. - [open note:2af4a7] K5 rel-kill-yale (MINOR): FIXED at its source with rel-kill (same sentence); MINOR, no page correction line. Left open with K4: the same old wording survives uncorrected in the there-here corpus copy (public/strata/there-here/index.html, research/lineage-edge-grammar/page.json). - [fixed] K6 body-cell (WRONG): FIXED: re-read LibreTexts 19.4 on 2026-10-18 (SDS single tails form spherical micelles, two-chain phospholipids bilayers). The dek now says the same packing rule curls soap into a ball and makes two-tailed lipids build the flat membrane, and the mismatch is why soap tears an enveloped virus apart; Corrected 2026-10-18 line added on the page and in the markdown. - [fixed] K7 body-hollow (MINOR): FIXED: 'build hollow balls' is now 'build tiny balls with an oily core'; MINOR, no page correction line. - [fixed] K8 body-onenumber (MINOR): FIXED: 'The whole story' is now 'Most of the story'; MINOR, no page correction line. - [fixed] K9 note-kill (MINOR): FIXED: the note now opens 'But soap is not mainly a germ killer, and it is not a universal solvent for germs'; MINOR, no page correction line. - [fixed] K10 note-kill-yale (MINOR): FIXED with note-kill (same sentence); MINOR, no page correction line. - [fixed] K11 note-polio (MINOR): FIXED: 'no lipid at all' is now 'no lipid envelope' (Chow et al. 1987 abstract re-read: myristic acid bound to a picornavirus capsid protein); MINOR, no page correction line. - [fixed] K12 note-rhino (MINOR): FIXED with note-polio (same phrase); MINOR, no page correction line. - [fixed] K13 js-cmc (MINOR): FIXED: the readout now says '(≈ 8.2 mM for SDS)'; MINOR, no page correction line. - [fixed] K14 js-gamma (MINOR): FIXED: the readout now marks the value illustrative, a straight line between the cited 72.8 and ~38; MINOR, no page correction line. - [fixed] K15 js-lift (MINOR): FIXED: the readout now says 'Grease lifted: N% (illustrative)'; MINOR, no page correction line. - [fixed] K16 js-cmcdial (MINOR): FIXED: the slider label is now 'How much soap' (no '× the CMC'); MINOR, no page correction line. - [declined] K17 ft-prov (MINOR): Lone MINOR (rule 4), metadata: the footer (index.html:252) names research/why-soap-cleans/facts.md as a repo path (not a link), and per the checker it is not served (404) while verify.mjs is served under /checks/. The fix is not a few words on this page: it needs facts.md published through the site's checks publishing (scripts/publish-checks.mjs) or the line reworded, so list it. - [fixed] K18 pk-everycell (MINOR): FIXED: packing.mjs:85 now says 'almost every living cell' (Wikipedia Archaea re-read: some archaea replace the bilayer with a monolayer); MINOR, no page correction line. - [fixed] K19 added-dial-sds (MINOR): FIXED: near the boundary (cylinder bin, p < 0.36) the dial now says 'self-assembles right at the sphere↔rod line' instead of naming a worm-like micelle, matching the table; MINOR, no page correction line.